Make kickass code less crappy

This does not actually fix anything; the issue was that the page was
gzipped. However, since the code was not checking anything, it took me
some time to realize it. We now have at least one "BrokenPageError"
check.

modules/kickass/browser.py

@@ -20,7 +20,7 @@
 
 from weboob.tools.browser import BaseBrowser
 
-from .pages.torrents import TorrentsPage, TorrentPage
+from .pages import TorrentsPage, TorrentPage
 
 
 __all__ = ['KickassBrowser']

modules/kickass/pages.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-# Copyright(C) 2010-2011 Julien Veyssier, Laurent Bachelier
+# Copyright(C) 2010-2012 Julien Veyssier, Laurent Bachelier
 #
 # This file is part of weboob.
 #
@@ -79,7 +79,7 @@ class TorrentPage(BasePage):
         for div in self.document.getiterator('div'):
             if div.attrib.get('id', '') == 'desc':
                 try:
-                    description = div.text_content()
+                    description = div.text_content().strip()
                 except UnicodeDecodeError:
                     description = 'Description with invalid UTF-8.'
             elif div.attrib.get('class', '') == 'seedBlock':
@@ -93,12 +93,13 @@ class TorrentPage(BasePage):
                 else:
                     leech = 0
 
-        for h in self.document.getiterator('h1'):
-            if h.attrib.get('class', '') == 'torrentName':
-                title = h.getchildren()[0].getchildren()[0].text
+        title = self.parser.select(self.document.getroot(),
+                'h1.torrentName span', 1)
+        title = title.text
 
         for a in self.document.getiterator('a'):
-            if ('Download' in a.attrib.get('title', '')) and ('torrent file' in a.attrib.get('title', '')):
+            if ('Download' in a.attrib.get('title', '')) \
+            and ('torrent file' in a.attrib.get('title', '')):
                 url = a.attrib.get('href', '')
 
         size = 0
@@ -106,7 +107,9 @@ class TorrentPage(BasePage):
         for span in self.document.getiterator('span'):
             # sometimes there are others span, this is not so sure but the size of the children list
             # is enough to know if this is the right span
-            if (span.attrib.get('class', '') == 'folder' or span.attrib.get('class', '') == 'folderopen') and len(span.getchildren()) > 2:
+            if (span.attrib.get('class', '') == 'folder' \
+                or span.attrib.get('class', '') == 'folderopen') \
+            and len(span.getchildren()) > 2:
                 size = span.getchildren()[1].tail
                 u = span.getchildren()[2].text
                 size = float(size.split(': ')[1].replace(',', '.'))